skip to main content

Title: Automated High‐Frequency Geomagnetic Disturbance Classifier: A Machine Learning Approach to Identifying Noise While Retaining High‐Frequency Components of the Geomagnetic Field

We present an automated method to identify high‐frequency geomagnetic disturbances in ground magnetometer data and classify the events by the source of the perturbations. We developed an algorithm to search for and identify changes in the surface magnetic field, dB/dt, with user‐specified amplitude and timescale. We used this algorithm to identify transient‐large‐amplitude (TLA) dB/dtevents that have timescale less than 60 s and amplitude >6 nT/s. Because these magnetic variations have similar amplitude and time characteristics to instrumental or man‐made noise, the algorithm identified a large number of noise‐type signatures as well as geophysical signatures. We manually classified these events by their sources (noise‐type or geophysical) and statistically characterized each type of event; the insights gained were used to more specifically define a TLA geophysical event and greatly reduce the number of noise‐type dB/dtidentified. Next, we implemented a support vector machine classification algorithm to classify the remaining events in order to further reduce the number of noise‐type dB/dtin the final data set. We examine the performance of our complete dB/dtsearch algorithm in widely used magnetometer databases and the effect of a common data processing technique on the results. The automated algorithm is a new technique to identify geomagnetic disturbances and instrumental or more » man‐made noise, enabling systematic identification and analysis of space weather related dB/dtevents and automated detection of magnetometer noise intervals in magnetic field databases.

« less
 ;  ;  ;  
Award ID(s):
2013648 1848724 1654044
Publication Date:
Journal Name:
Journal of Geophysical Research: Space Physics
DOI PREFIX: 10.1029
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    We present a characterization of transient‐large‐amplitude (TLA) geomagnetic disturbances that are relevant to geomagnetically induced currents (GIC). TLA events are defined as one or more short‐timescale (<60 s) dB/dt signature with magnitude ≥6 nT/s. The TLA events occurred at six stations of the Magnetometer Array for Cusp and Cleft Studies throughout 2015. A semi‐automated dB/dt search algorithm was developed to identify 38 TLA events in the ground magnetometer data. While TLA dB/dts do not drive GICs directly, we show that second‐timescale dB/dts often occur in relation to or within larger impulsive geomagnetic disturbances. Sudden commencements are not the main driver, rather the events are more likely to occur 30 min after a substorm onset or within a nighttime magnetic perturbation event. The characteristics of TLA events suggest localized ionospheric source currents that may play a key role in generating some extreme geomagnetic impulses that can lead to GICs.

  2. Abstract

    Nearly all studies of impulsive magnetic perturbation events (MPEs) with large magnetic field variability (dB/dt) that can produce dangerous geomagnetically induced currents (GICs) have used data from the Northern Hemisphere. Here we present details of four large‐amplitude MPE events (|ΔBx| > 900 nT and |dB/dt| > 10 nT/s in at least one component) observed between 2015 and 2018 in conjugate high‐latitude regions (65–80° corrected geomagnetic latitude), using magnetometer data from (1) Pangnirtung and Iqaluit in eastern Arctic Canada and the magnetically conjugate South Pole Station in Antarctica and (2) the Greenland West Coast Chain and two magnetically conjugate chains in Antarctica, AAL‐PIP and BAS LPM. From one to three different isolated MPEs localized in corrected geomagnetic latitude were observed during three premidnight events; many were simultaneous within 3 min in both hemispheres. Their conjugate latitudinal amplitude profiles, however, matched qualitatively at best. During an extended postmidnight interval, which we associate with an interval of omega bands, multiple highly localized MPEs occurred independently in time at each station in both hemispheres. These nighttime MPEs occurred under a wide range of geomagnetic conditions, but common to each was a negative interplanetary magnetic fieldBzthat exhibited at least a modest increase at or near the time of the event. Amore »comparison of perturbation amplitudes to modeled ionospheric conductances in conjugate hemispheres clearly favored a current generator model over a voltage generator model for three of the four events; neither model provided a good fit for the premidnight event that occurred near vernal equinox.

    « less
  3. The historical record indicates the possibility of intense coronal mass ejections (CMEs). Energized particles and magnetic fields ejected by coronal mass ejections (CMEs) towards the Earth may disrupt the Earth’s magnetosphere and generate a geomagnetic storm. During a geomagnetic storm, the induced geoelectric field can drive geomagnetically-induced currents (GICs) that flow through ground-based conductors. These GICs have the potential to damage high voltage power transmission systems and cause blackouts. As part of the NSF-funded Comprehensive Hazard Analysis for Resilience to Geomagnetic Extreme Disturbances (CHARGED) project, a solar-wind-to-lithosphere numerical model of the geoelectric field is being developed. The purpose of this new tool is to drive a new generation of GIC forecasting. As a part of that work, Maxwell’s equations, finite-difference time-domain (FDTD) models of the last stage of the Sun-to-Earth propagation path is being coupled to output generated by the Block Adaptive Tree Solarwind Roe-type Upwind Scheme (BATS-R-US) magnetohydrodynamics model and the Ridley Ionosphere Model (RIM) of ionospheric dynamics. Specifically, three-dimensional (3-D) BATS-R-US and RIM-predicted ionospheric currents occurring in the lower ionosphere during and around the time of the March 17, 2015 storm are modeled in 3-D FDTD models of North America. These models start at a depth of 150more »km, and they account for ionospheric currents occurring up to an altitude of 115 km. The resolution of the FDTD models is 22 km (East-West) x 11 km (North-South) x 5 km (radially), and they account for 3-D lithosphere conductivities provided by the U.S. Geological Survey. The FDTD-calculated results are compared with surface magnetic fields measured in the region by SuperMAG and INTERMAGNET magnetometers. The FDTD results are also compared with virtual magnetometer data, which calculates the perturbation of the surface magnetic field using output from the BATS-R-US magnetohydrodynamics model. Comparison plots and an analysis of the results will be provided.« less
  4. Abstract

    The rapid changes of magnetic fields associated with large, isolated magnetic perturbations with amplitudes |ΔB| of hundreds of nanotesla and 5‐ to 10‐min periods can induce bursts of geomagnetically induced currents that can harm technological systems. This paper presents statistical summaries of the characteristics of nightside magnetic perturbation events observed in Eastern Arctic Canada from 2014 through 2017 using data from stations that are part of four magnetometer arrays: MACCS, AUTUMNX, CANMOS, and CARISMA, covering a range of magnetic latitudes from 68 to 78°. Most but not all of the magnetic perturbation events were associated with substorms: roughly two thirds occurred between 5 and 30 min after onset. The association of intense nighttime magnetic perturbation events with magnetic storms was significantly reduced at latitudes above 73°, presumably above the nominal auroral oval. A superposed epoch study of 21 strong events at Cape Dorset showed that the largest |dB/dt| values appeared within an ~275‐km radius that was associated with a region of shear between upward and downward field‐aligned currents. The statistical distributions of impulse amplitudes of both |ΔB| and |dB/dt| fit well the log‐normal distribution at all stations. The |ΔB| distributions are similar over the magnetic latitude range studied, butmore »the kurtosis and skewness of the |dB/dt| distributions show a slight increase with latitude. Knowledge of the statistical characteristics of these events has enabled us to estimate the occurrence probability of extreme impulsive disturbances using the approximation of a log‐normal distribution.

    « less
  5. Abstract

    We must be able to predict and mitigate against geomagnetically induced current (GIC) effects to minimize socio‐economic impacts. This study employs the space weather modeling framework (SWMF) to model the geomagnetic response over Fennoscandia to the September 7–8, 2017 event. Of key importance to this study is the effects of spatial resolution in terms of regional forecasts and improved GIC modeling results. Therefore, we ran the model at comparatively low, medium, and high spatial resolutions. The virtual magnetometers from each model run are compared with observations from the IMAGE magnetometer network across various latitudes and over regional‐scales. The virtual magnetometer data from the SWMF are coupled with a local ground conductivity model which is used to calculate the geoelectric field and estimate GICs in a Finnish natural gas pipeline. This investigation has lead to several important results in which higher resolution yielded: (1) more realistic amplitudes and timings of GICs, (2) higher amplitude geomagnetic disturbances across latitudes, and (3) increased regional variations in terms of differences between stations. Despite this, substorms remain a significant challenge to surface magnetic field prediction from global magnetohydrodynamic modeling. For example, in the presence of multiple large substorms, the associated large‐amplitude depressions were notmore »captured, which caused the largest model‐data deviations. The results from this work are of key importance to both modelers and space weather operators. Particularly when the goal is to obtain improved regional forecasts of geomagnetic disturbances and/or more realistic estimates of the geoelectric field.

    « less